INTRODUCTORY ESSAY. 95 Bengal, the west coast of the Madras peninsula, and of Cey- lon, contrast strongly with the drier parts of the inter tropical zone, and still more so with the loosely-timbered districts of Central India, and of the hase of the western Himalaya. The absolutely sterile deserts are confined to the extensive plains, which arc all cut off from the rains by being placed to leeward of mountain-ranges, or by other causes. Thcrfc are hence in India no vast plains clothed \\itli gigantic timber- trees, such .as cover immense areas of the American tropic^; and even ihe valleys of the great Indian rivers, tire Ganges, Nerbada, etc., are nowhere heavily timbered, -but are gene- rally absolutely destitute of forest, and extremely populous and highly cultivated*. The tropical forests of India may be divided into those whicli inhabit perennially humid districts, and those which arc confined to regions presenting contrasted seasons, of sum- mer rain and winter drought. The jjerennially humid forests arc uniformly characterized by the prevalence of Ferns, and, at elevations below 5000- 7000 feet, by the immense number of epiphytal Orchidetc, Orontiacea, and Scitawi/ti'M : they contain a far greater amount of species than the drier forests, and arc i'urther cha- racterized by Ziiiyibtraceai, XyritieM) Palms, Pardaiuw*, Dro- ctena, Piper, Chloran&hus, Urticacea (especially Fici)t Araliacea, Apocyneai^ shrubby Ruhutctw, Garciniacea, Ano/iacea, Nutmegs, and Dij The drier tropical forests of the regions with contrasted seasons, arc much modified in luxuriance and extension by the winter cold in those extratropical latitudes over which they spread. In the chapter upon the meteorology of India, it is shown that though the summer heat scarcely decrease* * It is a much discussed question in India, whethcT the Gangctic plain waa ever covered with forest s the best authorities consider that it never WM so; but there are others who hold the contrary opinion, and aver that the deutruction of the timber hus produced a great change in the donate. The absence of vege- table remains in the alluvium appears unfavourable to the latter opinion.